No description
Find a file
2022-07-28 08:42:28 -05:00
.github use jruby 9.3.6 2022-07-26 09:57:40 -05:00
lib standardize Name tag, implement tag_vm_user method 2022-07-28 08:42:28 -05:00
spec standardize Name tag, implement tag_vm_user method 2022-07-28 08:42:28 -05:00
.gitignore rename from AWS to EC2 for consistency 2022-07-07 08:49:46 -05:00
.jrubyrc Added aws dependency and renamed directories 2022-07-06 13:41:54 -05:00
.rubocop.yml fix rubocop offenses 2022-07-06 14:49:04 -05:00
CODEOWNERS CODEOWNERS by slack-gitbot 2022-06-09 13:56:54 -04:00
Gemfile Added aws dependency and renamed directories 2022-07-06 13:41:54 -05:00
LICENSE Added aws dependency and renamed directories 2022-07-06 13:41:54 -05:00
Rakefile Added aws dependency and renamed directories 2022-07-06 13:41:54 -05:00
README.md integrtae GCP Cloud DNS from the gce provider 2022-07-27 10:19:22 -05:00
vmpooler-provider-ec2.gemspec integrtae GCP Cloud DNS from the gce provider 2022-07-27 10:19:22 -05:00
vmpooler.yaml.example integrtae GCP Cloud DNS from the gce provider 2022-07-27 10:19:22 -05:00

vmpooler-provider-aws

This is a provider for VMPooler allows using aws to create instances, disks, snapshots, or destroy instances for specific pools.

Usage

Include this gem in the same Gemfile that you use to install VMPooler itself and then define one or more pools with the provider key set to aws. VMPooler will take care of the rest. See what configuration is needed for this provider in the example file.

Examples of deploying VMPooler with extra providers can be found in the puppetlabs/vmpooler-deployment repository.

aws authorization is handled via two required ENV vars

  1. ABS_AWS_ACCESS_KEY
  2. ABS_AWS_SECRET_KEY

Provisioning the new nodes

When you add the pool config provision: true to a pool, the new VMs will also get initialized with extra steps to setup the sshd config via NET:SSH These steps expect two environment vars

  1. ROOT_KEYS_SCRIPT: (optional) the URI location of a script (eg https in github) that will be run to setup keys. If not set, this will be skipped
  2. AWS_KEY_FILE_LOCATION: (required) the location on local disk where the ssh key resides for VMPooler to connect via SSH to the EC2 node

DNS

AWS will setup a private ip and private dns hostname for the VM once running. Optionally we can setup a human readable DNS entry to resolve the VMPooler provider spicy-proton fqdn

DNS is integrated via Google's CloudDNS service. GCE authorization is handled via a service account (or personal account) private key (json format) and can be configured via

  1. GOOGLE_APPLICATION_CREDENTIALS environment variable eg GOOGLE_APPLICATION_CREDENTIALS=/my/home/directory/my_account_key.json

Provider config needed:

  1. domain
  2. project
  3. dns_zone_resource_name (see the example yaml file)

An A record is then created in that zone upon instance creation with the VM's internal IP, and deleted when the instance is destroyed.

Labels

This provider adds tags to all resources that are managed

resource labels note
instance vm=$vm_name, pool=$pool_name for example vm=foo-bar, pool=pool1
disk vm=$vm_name, pool=$pool_name for example vm=foo-bar and pool=pool1
snapshot snapshot_name=$snapshot_name, vm=$vm_name, pool=$pool_name for example snapshot_name=snap1, vm=foo-bar, pool=pool1

Also see the usage of vmpooler's optional purge_unconfigured_resources, which is used to delete any resource found that do not have the pool label, and can be configured to allow a specific list of unconfigured pool names.

Pre-requisite

  • An IAM user must exist in the target AWS account with permissions to create, delete vms etc
  • if using DNS see section above, and a service account with permissions to change Cloud DNS need to exist

License

vmpooler-provider-aws is distributed under the Apache License, Version 2.0. See the LICENSE file for more details.